Original Report: Laboratory Investigation American
Journal of
Nephrology
Received: October 13, 2014 Accepted: December 10, 2014 Published online: January 23, 2015
Am J Nephrol 2015;41:7–15 DOI: 10.1159/000371557
Estimated GFR Is Biased by Non-Traditional Cardiovascular Risk Factors Toralf Melsom a, c Ole Martin Fuskevåg b Ulla Dorte Mathisen a, c Harald Strand b Jørgen Schei a, c Trond Jenssen c, d Marit Solbu a, c Bjørn Odvar Eriksen a, c
Departments of a Nephrology and b Laboratory Medicine, University Hospital of North Norway, UNN, Tromsø; c Metabolic and Renal Research Group, UiT The Arctic University of Norway, and d Department of Nephrology, Oslo University Hospital, Rikshospitalet, Oslo, Norway
Abstract Background: Estimated glomerular filtration rate (eGFR) based on either cystatin C or creatinine performs similarly in estimating measured GFR, but associate differently with cardiovascular disease (CVD) and mortality. This could be due to confounding by non-GFR-related traits associated with cystatin C and creatinine levels. We investigated non-GFRrelated associations between eGFR and two types of nontraditional risk factors for CVD and death: L-arginine/dimethylarginine metabolism and insulin resistance. Methods: GFR was measured via iohexol clearance in a cross-sectional study of 1,624 middle-aged persons from the general population without CVD, diabetes or chronic kidney disease. The dimethylarginines were measured using liquid chromatography tandem mass spectrometry (LC-MSMS). Insulin resistance was determined by the homeostasis model assessment (HOMA-IR). Results: Asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA), the L-arginine/ADMA ratio and insulin resistance were associated with creatinine-based eGFR after accounting for measured GFR in
© 2015 S. Karger AG, Basel 0250–8095/15/0411–0007$39.50/0 E-Mail [email protected] www.karger.com/ajn
multivariable adjusted analyses. The cystatin C-based eGFR showed a similar residual association with SDMA; an oppositely directed, borderline significant association with ADMA; and a stronger residual association with insulin resistance compared with eGFR based on creatinine. Conclusion: Both creatinine- and cystatin C-based eGFR are influenced by nontraditional risk factors, which may bias risk prediction by eGFR in longitudinal studies. © 2015 S. Karger AG, Basel
Introduction
An estimated glomerular filtration rate (eGFR) below 60 ml/min/1.73 m2 is an important risk factor for endstage renal disease, cardiovascular disease (CVD) and death. Epidemiological studies have found that eGFR measurements based on cystatin C (eGFRcys) show a stronger association with renal and CVD outcomes than GFR estimated according to the creatinine level (eGFRcre) [1, 2]. Because there is little evidence indicating that cystatin C is more effective for estimating GFR compared to creatinine, this superior risk prediction by eGFRcys may be caused by the confounding of non-GFR determinants of cystatin C and creatinine [3, 4]. Several studies have shown eGFR to be influenced by traditional Toralf Melsom, MD Section of Nephrology University Hospital of North Norway NO–9038 Tromsø (Norway) E-Mail toralf.melsom @ unn.no
Downloaded by: Kainan University 203.64.11.45 - 2/22/2015 12:07:58 PM
Key Words GFR estimation · L-arginine · Dimethylarginines · Insulin resistance · Cardiovascular disease risk factors · Epidemiology · Creatinine and cystatin C
Subjects and Methods Participants The RENIS-T6 is a substudy of the sixth population-based Tromsø study (Tromsø 6). Tromsø 6 included an age-stratified representative sample of 12,984 inhabitants of the municipality of Tromsø in Northern Norway [24]. Forty percent of all inhabitants between 50 and 59 years of age and all inhabitants between 60 and
8
Am J Nephrol 2015;41:7–15 DOI: 10.1159/000371557
62 years of age were invited to participate in Tromsø 6. In these age groups, 3,564 subjects (65%) completed the main part of the Tromsø 6 survey; of these, we excluded 739 who reported previous myocardial infarction, angina pectoris, stroke, diabetes mellitus or renal disease (fig. 1). The remaining 2,825 individuals were invited to participate in RENIS-T6, and 2,107 (75%) responded positively. A detailed description of the study participants and methods of the RENIS-T6 has been published elsewhere [25]. Briefly, 77 individuals were excluded because of allergies to contrast media and 48 individuals did not appear at their appointments. Among the remaining 1,982 individuals, we included 1,632 individuals according to a predetermined target. Five participants were excluded because of technical failure in the GFR measurements, leaving 1,627 persons in the RENIS-T6 cohort (fig. 1). The characteristics of the RENIS-T6 cohort were comparable to the total group of eligible subjects (n = 2,825) [26]. In the present investigation, we excluded 3 individuals because of methylarginine measurement failure. The study adhered to the Declaration of Helsinki and was approved by the Regional Ethics Committee of Northern Norway. All participants provided informed written consent. Measurements All study participants met between 8:00 and 10:00 a.m. in the Clinical Research Unit at the University Hospital of Northern Norway after an overnight fast. Blood pressure and blood samples were obtained after the participants had been resting for at least at least 5 min. Serum samples for glucose, creatinine, triglycerides and cholesterol were measured the same day. Serum samples used for measuring the levels of ADMA, SDMA, L-arginine and insulin were stored at –80 ° C and thawed at the time of analysis. Additional information about CVD risk factors was obtained using a questionnaire. Smoking status was divided into current smokers and non-smokers. A family history of early myocardial infarction was defined as a first-degree relative with myocardial infarction before the age of 60.
Iohexol Clearance GFR was measured as the single-sample plasma clearance of iohexol, as previously described in detail [25]. This method has been validated against gold standard methods [27, 28]. Briefly, 5 ml of iohexol (Omnipaque, 300 mg I/ml, Amersham Health, London, UK) was injected intravenously. The exact time for measuring the iohexol concentration was calculated with Jacobsson’s method based on the eGFRcre [29]. The iohexol concentration was measured using high-performance liquid chromatography. The between-run coefficient of variation (CV) during the study period was 3.0%. GFR was calculated using the formulas described by Jacobsson [29]. L-Arginine, ADMA and SDMA Measurement ADMA, SDMA and L-arginine were analyzed by LC-MSMS using the Waters AcquityTM UPLC system with an auto sampler and a binary solvent delivery system (Waters, Milford, Mass., USA) interfaced to the Waters Xevo TQ-S benchtop tandem quadrupole mass spectrometer (Waters, Manchester, UK). Chromatography was performed on a 1.7 μm, 2.1 × 100 mm Waters Acquity UPLC BEH Amide column maintained at 50 ° C. The column was eluted isocratically using 10 mM ammonium formate with 0.1% formic acid in water-acetonitrile at 0.5 ml/min. Reference standards for ADMA, SDMA and L-arginine were purchased from Sig
Melsom/Fuskevåg/Mathisen/Strand/ Schei/Jenssen/Solbu/Eriksen
Downloaded by: Kainan University 203.64.11.45 - 2/22/2015 12:07:58 PM
CVD risk factors independently of the measured GFR (mGFR). In particular, cystatin C seems to be associated with obesity, smoking, HDL-cholesterol, triglycerides, hypertension and C-reactive protein (CRP) [5–7]. These findings indicate that the estimated risk associated with reduced eGFR in longitudinal studies may be confounded, particularly when eGFRcys is used. However, because most epidemiological studies reduce confounding by adjusting their survival analysis for these traditional CVD risk factors, this may not explain the large difference between eGFRcys and eGFRcre risk estimates. Novel CVD risk factors are more difficult to measure in population studies and therefore not commonly used for adjustment in survival analysis. Accordingly, confounding by these risk factors may contribute to different risk predictions generated according to creatinine- and cystatin C-based eGFR. In this study, we investigated the association between eGFR based on the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations [4] and two types of variables representing nontraditional CVD risk factors: arginine/dimethylarginine metabolism and insulin resistance. Asymmetric dimethylarginine (ADMA), its isomer symmetric dimethylarginine (SDMA), and the L-arginine/ADMA ratio have recently emerged as strong and independent risk factors for CVD and death, both in the general population and in high-risk patients [8–13]. Insulin resistance (IR) also predicts CVD and death in community cohorts [14–18], and IR and dimethylarginines are hypothesized to interact in a vicious cycle leading to endothelial dysfunction and vascular disease [19–22]. In the Renal Iohexol Clearance Survey in Tromsø 6 (RENIS-T6), we measured GFR by iohexol clearance as well as the levels of L-arginine, ADMA, SDMA and IR in a middle-aged cohort from the general population. Using a cross-sectional design, we aimed to explore whether these novel risk factors are determinants of eGFR, after accounting for mGFR. In addition, we sought to study the relationship between mGFR and dimethylarginines. The relationship between mGFR and IR has been reported previously [23].
n = 5,464 Invited to participate the Tromsø 6 study and aged 50 to 62 years
n = 3,564 (65%) Completed the main Tromsø 6 study
n = 739 Reported a previous myocardial infarction, angina pectoris, stroke, diabetes mellitus, or any renal disease except for urinary tract infection
n = 2,825 Invited to participate in RENIS-T6
n = 2,107 (74%) Responded
n = 1,982 Eligible for inclusion
n = 125 Excluded because of allergy to contrast media, iodine, or latex, or for other reasons. Includes 48 subjects who withdrew from the study
n = 1,632 Investigated in RENIS-T6 according to a predetermined target
n=5 Technical failure in the iohexol-clearance measurements
n = 1,627 The RENIS-T6 cohort n=3 Technical failure in the dimethylarginine measurements n = 1,624 Present study population
Fig. 1. Flowchart of the study population.
ma-Aldrich (St. Louis, Mo., USA). The stable isotope-labeled internal standards D7-ADMA hydrochloride and L-arginine-d7 hydrochloride were obtained from Cambridge Isotope Laboratories (Andover, Mass., USA). All other chemicals were of an analytical grade or better. Standard samples, quality control samples and 50 μl of unknown serum samples were mixed with 450 μl of the internal standard/precipitation solution in a Waters 96-well PP sample collection plate (Waters, Milford, Mass., USA). The plate was sealed and placed at 4 ° C for 30 min before centrifugation at 300 × g for 10 min. The plate was then put back on the liquid handler, and 50 μl of the clear supernatant was transferred to a second Waters 96-well collection plate and diluted with 200 μl water, sealed and placed in the UPLC autosampler for analysis. The between-day CV was
Journal of
Nephrology
Received: October 13, 2014 Accepted: December 10, 2014 Published online: January 23, 2015
Am J Nephrol 2015;41:7–15 DOI: 10.1159/000371557
Estimated GFR Is Biased by Non-Traditional Cardiovascular Risk Factors Toralf Melsom a, c Ole Martin Fuskevåg b Ulla Dorte Mathisen a, c Harald Strand b Jørgen Schei a, c Trond Jenssen c, d Marit Solbu a, c Bjørn Odvar Eriksen a, c
Departments of a Nephrology and b Laboratory Medicine, University Hospital of North Norway, UNN, Tromsø; c Metabolic and Renal Research Group, UiT The Arctic University of Norway, and d Department of Nephrology, Oslo University Hospital, Rikshospitalet, Oslo, Norway
Abstract Background: Estimated glomerular filtration rate (eGFR) based on either cystatin C or creatinine performs similarly in estimating measured GFR, but associate differently with cardiovascular disease (CVD) and mortality. This could be due to confounding by non-GFR-related traits associated with cystatin C and creatinine levels. We investigated non-GFRrelated associations between eGFR and two types of nontraditional risk factors for CVD and death: L-arginine/dimethylarginine metabolism and insulin resistance. Methods: GFR was measured via iohexol clearance in a cross-sectional study of 1,624 middle-aged persons from the general population without CVD, diabetes or chronic kidney disease. The dimethylarginines were measured using liquid chromatography tandem mass spectrometry (LC-MSMS). Insulin resistance was determined by the homeostasis model assessment (HOMA-IR). Results: Asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA), the L-arginine/ADMA ratio and insulin resistance were associated with creatinine-based eGFR after accounting for measured GFR in
© 2015 S. Karger AG, Basel 0250–8095/15/0411–0007$39.50/0 E-Mail [email protected] www.karger.com/ajn
multivariable adjusted analyses. The cystatin C-based eGFR showed a similar residual association with SDMA; an oppositely directed, borderline significant association with ADMA; and a stronger residual association with insulin resistance compared with eGFR based on creatinine. Conclusion: Both creatinine- and cystatin C-based eGFR are influenced by nontraditional risk factors, which may bias risk prediction by eGFR in longitudinal studies. © 2015 S. Karger AG, Basel
Introduction
An estimated glomerular filtration rate (eGFR) below 60 ml/min/1.73 m2 is an important risk factor for endstage renal disease, cardiovascular disease (CVD) and death. Epidemiological studies have found that eGFR measurements based on cystatin C (eGFRcys) show a stronger association with renal and CVD outcomes than GFR estimated according to the creatinine level (eGFRcre) [1, 2]. Because there is little evidence indicating that cystatin C is more effective for estimating GFR compared to creatinine, this superior risk prediction by eGFRcys may be caused by the confounding of non-GFR determinants of cystatin C and creatinine [3, 4]. Several studies have shown eGFR to be influenced by traditional Toralf Melsom, MD Section of Nephrology University Hospital of North Norway NO–9038 Tromsø (Norway) E-Mail toralf.melsom @ unn.no
Downloaded by: Kainan University 203.64.11.45 - 2/22/2015 12:07:58 PM
Key Words GFR estimation · L-arginine · Dimethylarginines · Insulin resistance · Cardiovascular disease risk factors · Epidemiology · Creatinine and cystatin C
Subjects and Methods Participants The RENIS-T6 is a substudy of the sixth population-based Tromsø study (Tromsø 6). Tromsø 6 included an age-stratified representative sample of 12,984 inhabitants of the municipality of Tromsø in Northern Norway [24]. Forty percent of all inhabitants between 50 and 59 years of age and all inhabitants between 60 and
8
Am J Nephrol 2015;41:7–15 DOI: 10.1159/000371557
62 years of age were invited to participate in Tromsø 6. In these age groups, 3,564 subjects (65%) completed the main part of the Tromsø 6 survey; of these, we excluded 739 who reported previous myocardial infarction, angina pectoris, stroke, diabetes mellitus or renal disease (fig. 1). The remaining 2,825 individuals were invited to participate in RENIS-T6, and 2,107 (75%) responded positively. A detailed description of the study participants and methods of the RENIS-T6 has been published elsewhere [25]. Briefly, 77 individuals were excluded because of allergies to contrast media and 48 individuals did not appear at their appointments. Among the remaining 1,982 individuals, we included 1,632 individuals according to a predetermined target. Five participants were excluded because of technical failure in the GFR measurements, leaving 1,627 persons in the RENIS-T6 cohort (fig. 1). The characteristics of the RENIS-T6 cohort were comparable to the total group of eligible subjects (n = 2,825) [26]. In the present investigation, we excluded 3 individuals because of methylarginine measurement failure. The study adhered to the Declaration of Helsinki and was approved by the Regional Ethics Committee of Northern Norway. All participants provided informed written consent. Measurements All study participants met between 8:00 and 10:00 a.m. in the Clinical Research Unit at the University Hospital of Northern Norway after an overnight fast. Blood pressure and blood samples were obtained after the participants had been resting for at least at least 5 min. Serum samples for glucose, creatinine, triglycerides and cholesterol were measured the same day. Serum samples used for measuring the levels of ADMA, SDMA, L-arginine and insulin were stored at –80 ° C and thawed at the time of analysis. Additional information about CVD risk factors was obtained using a questionnaire. Smoking status was divided into current smokers and non-smokers. A family history of early myocardial infarction was defined as a first-degree relative with myocardial infarction before the age of 60.
Iohexol Clearance GFR was measured as the single-sample plasma clearance of iohexol, as previously described in detail [25]. This method has been validated against gold standard methods [27, 28]. Briefly, 5 ml of iohexol (Omnipaque, 300 mg I/ml, Amersham Health, London, UK) was injected intravenously. The exact time for measuring the iohexol concentration was calculated with Jacobsson’s method based on the eGFRcre [29]. The iohexol concentration was measured using high-performance liquid chromatography. The between-run coefficient of variation (CV) during the study period was 3.0%. GFR was calculated using the formulas described by Jacobsson [29]. L-Arginine, ADMA and SDMA Measurement ADMA, SDMA and L-arginine were analyzed by LC-MSMS using the Waters AcquityTM UPLC system with an auto sampler and a binary solvent delivery system (Waters, Milford, Mass., USA) interfaced to the Waters Xevo TQ-S benchtop tandem quadrupole mass spectrometer (Waters, Manchester, UK). Chromatography was performed on a 1.7 μm, 2.1 × 100 mm Waters Acquity UPLC BEH Amide column maintained at 50 ° C. The column was eluted isocratically using 10 mM ammonium formate with 0.1% formic acid in water-acetonitrile at 0.5 ml/min. Reference standards for ADMA, SDMA and L-arginine were purchased from Sig
Melsom/Fuskevåg/Mathisen/Strand/ Schei/Jenssen/Solbu/Eriksen
Downloaded by: Kainan University 203.64.11.45 - 2/22/2015 12:07:58 PM
CVD risk factors independently of the measured GFR (mGFR). In particular, cystatin C seems to be associated with obesity, smoking, HDL-cholesterol, triglycerides, hypertension and C-reactive protein (CRP) [5–7]. These findings indicate that the estimated risk associated with reduced eGFR in longitudinal studies may be confounded, particularly when eGFRcys is used. However, because most epidemiological studies reduce confounding by adjusting their survival analysis for these traditional CVD risk factors, this may not explain the large difference between eGFRcys and eGFRcre risk estimates. Novel CVD risk factors are more difficult to measure in population studies and therefore not commonly used for adjustment in survival analysis. Accordingly, confounding by these risk factors may contribute to different risk predictions generated according to creatinine- and cystatin C-based eGFR. In this study, we investigated the association between eGFR based on the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations [4] and two types of variables representing nontraditional CVD risk factors: arginine/dimethylarginine metabolism and insulin resistance. Asymmetric dimethylarginine (ADMA), its isomer symmetric dimethylarginine (SDMA), and the L-arginine/ADMA ratio have recently emerged as strong and independent risk factors for CVD and death, both in the general population and in high-risk patients [8–13]. Insulin resistance (IR) also predicts CVD and death in community cohorts [14–18], and IR and dimethylarginines are hypothesized to interact in a vicious cycle leading to endothelial dysfunction and vascular disease [19–22]. In the Renal Iohexol Clearance Survey in Tromsø 6 (RENIS-T6), we measured GFR by iohexol clearance as well as the levels of L-arginine, ADMA, SDMA and IR in a middle-aged cohort from the general population. Using a cross-sectional design, we aimed to explore whether these novel risk factors are determinants of eGFR, after accounting for mGFR. In addition, we sought to study the relationship between mGFR and dimethylarginines. The relationship between mGFR and IR has been reported previously [23].
n = 5,464 Invited to participate the Tromsø 6 study and aged 50 to 62 years
n = 3,564 (65%) Completed the main Tromsø 6 study
n = 739 Reported a previous myocardial infarction, angina pectoris, stroke, diabetes mellitus, or any renal disease except for urinary tract infection
n = 2,825 Invited to participate in RENIS-T6
n = 2,107 (74%) Responded
n = 1,982 Eligible for inclusion
n = 125 Excluded because of allergy to contrast media, iodine, or latex, or for other reasons. Includes 48 subjects who withdrew from the study
n = 1,632 Investigated in RENIS-T6 according to a predetermined target
n=5 Technical failure in the iohexol-clearance measurements
n = 1,627 The RENIS-T6 cohort n=3 Technical failure in the dimethylarginine measurements n = 1,624 Present study population
Fig. 1. Flowchart of the study population.
ma-Aldrich (St. Louis, Mo., USA). The stable isotope-labeled internal standards D7-ADMA hydrochloride and L-arginine-d7 hydrochloride were obtained from Cambridge Isotope Laboratories (Andover, Mass., USA). All other chemicals were of an analytical grade or better. Standard samples, quality control samples and 50 μl of unknown serum samples were mixed with 450 μl of the internal standard/precipitation solution in a Waters 96-well PP sample collection plate (Waters, Milford, Mass., USA). The plate was sealed and placed at 4 ° C for 30 min before centrifugation at 300 × g for 10 min. The plate was then put back on the liquid handler, and 50 μl of the clear supernatant was transferred to a second Waters 96-well collection plate and diluted with 200 μl water, sealed and placed in the UPLC autosampler for analysis. The between-day CV was
